https://bugs.kde.org/show_bug.cgi?id=363522
Bug ID: 363522 Summary: import: use UUID from XMP sidecar / metadata instead of creating another one Product: digikam Version: 4.14.0 Platform: Ubuntu Packages OS: Linux Status: UNCONFIRMED Severity: normal Priority: NOR Component: Import Assignee: digikam-de...@kde.org Reporter: jens-bugs.kde....@spamfreemail.de When I import images that have existing XMP sidecar files Digikam reads those files and uses the metadata in them. But it creates its own UUID for the database although there already is one supplied in the XMP sidecar, as <rdf:Description ... ... digiKam:ImageUniqueID="QXbwXF%yQEqRFBbG+GkfyA" ... > This is important if you reimport files again so they are uniquely identifiable even from an external script. The UUID is the only *safe* way of uniquely identifying an image after the import (after all, it might have been renamed, moved, etc). I suppose the UUID really is just a unique random string of any length (right?). Duplicate UUIDs, if present, should be handled just like duplicate files, by prompting the user to replace the file or create a new UUID when first reading the image file's metadata. But in reality this should never happen - it is just important that Digikam doesn't crash with a duplicate UUID. Reproducible: Always -- You are receiving this mail because: You are watching all bug changes.